In their Comment, Messemaker et al. discuss genetic data that show that block 2 within the TRAF1-C5 locus is a risk for rheumatoid arthritis (RA), while block 3 is not associated. We acknowledge these data, which illustrate that, when taken in isolation, block 3 does not impact disease risk. It is clear that the genetic data are robust (1); however, there is no question that the rs17611 single nucleotide polymorphism (SNP) in C5 results in an amino acid change in C5 that alters its rate of cleavage by elastase: an enzyme present at high levels in neutrophil-driven or neutrophil-associated diseases such as RA. We believe that it would be important and interesting to further analyze whether this functional polymorphism in block 3, which affects the proinflammatory capacity of C5, has any impact on the risk haplotype in block 2. While genome-wide association studies have identified a number of independent secondary association signals, these are not exclusive (2); the strong functional data that we demonstrate with the C5-V802I variant suggest that potential interactions should be specifically tested in future studies to confirm or exclude them.
In our article (3), we reference other papers where the specific rs17611 SNP has been linked to different diseases. Chai et al. (4) show an association (p < 0.007) of rs17611 (and its linked set) with periodontal disease; Woehrl et al. (5) show an association (p < 0.002) of rs17611 with outcome in pneumococcal meningitis; Hoke et al. (6) show an association (p < 0.01) of rs17611 with adverse cardiovascular outcome; and Greisenegger et al. (7) show an association (p < 0.005) of rs17611 with risk for ischemic stroke. Together, these linkages provided a strong rationale to explore the functional consequences of the rs17611 SNP on C5 activities, the major focus of our paper.
First, we show that the rs17611 SNP is associated with clear differences in C5 turnover and elevated levels of the proinflammatory product C5a in healthy individuals and RA patients; our finding warrants further investigation of the impact of this polymorphism on risk associated with block 2. We go on to define the mechanism by which the single amino acid change provokes these differences. The genetic data do not detract from the importance of our demonstration that the polymorphism impacts C5 turnover and increases plasma levels of the proinflammatory molecule C5a, and of our description of the atypical route by which this is achieved (3).
